FACTORS INFLUENCING WOMEN’S CHOICE OF PLACE OF DELIVERY IN RURAL MALAWI -
An explorative study
By Line Seljeskog, 1 Johanne Sundby 2and Jane Chimango3
1: Medical student, University of Oslo, Norway.
Correspondence: E-mail: [email protected]. Private address: Trudvangveien 21, 0363 Oslo, Norway.
2: MD, PhD, Professor, Section for International Health, University of Oslo.
Correspondence: Section for International Health, University of Oslo, Post box 1130 Blindern, 0318 Oslo, Norway. E-mail: [email protected]
3: Community and Mental Health Nursing Departement, Kamuzu College of Nursing, University of Malawi.
KEY WORDS: Maternal health Quality of care Place of delivery
ABSTRACT
This explorative study was conducted in the Mangochi area in Malawi with the aim of
investigating the individual, community and health facility level factors influencing women’s choice of place of delivery. In depth interviews and non-participating observation were the methods used. Three major issues were revealed. First, sub optimal quality of care including communication, attitudes and cooperation within the health care system was identified as a main factor. Secondly, cultural aspects such as influence from decision makers, perceptions of danger signs and traditional views on pregnancy and delivery were important. Finally, an unsatisfactory availability to skilled delivery care in terms of distance, transport and costs was shown. We conclude that the barriers to use of professional obstetric care in Malawi partly can be attributed to the health care system itself, and that a more individualized maternity care is needed.
INTRODUCTION
Annually, more than 500,000 maternal deaths occur worldwide, the majority equally divided between Africa and Asia.1 Less than 1% of the pregnancy-related deaths occur in the more developed parts of the world, making maternal mortality the health indicator showing the greatest disparity between developing and developed countries.2Despite the enhanced focus and awareness over the past decades, the situation in the poorest countries has not improved, and maternal mortality reduction is one of the explicit health millennium development goals.3
In Malawi, estimates of the maternal mortality ratio suggests that it has actually increased over the past years, and considerably so, from 620 per 100,000 live births in 1992 to almost 1,120 in 2000.4 The lifetime risk of maternal death is 1 in 7, which contrasts to the world average at 1 in 74 and especially to the lifetime risk in industrial countries at 1 in 4,085.1 Even though the accuracy of the numbers is uncertain, the difference is striking.
Access to professional health care during delivery is considered to be critical for maternal mortality reduction. According to Buttiëns, Marchal and De Brouwere, 5 the rationale for this is grounded in three observations. First, early identification and proper management of obstetric complications is fundamental for life saving. The majority of deaths occur during delivery. Moreover, the antenatal screening has such a low predictive value of identifying risk cases, that the presence of a skilled attendant at delivery more likely reduces mortality.
Secondly, the common factor among countries managing to reduce pregnancy-related deaths has been the presence of skilled providers attending most of the deliveries. Finally,
ineffectiveness of training traditional birth attendants has drawn attention towards training of professional personnel. The proportion of births conducted by a skilled attendant has become an indicator for monitoring progress towards reduction of maternal mortality. 6
In Malawi, the antenatal care coverage is almost complete (>95%). However, this high attendance does not translate into high institutional delivery rates, with only 41% of the deliveries conducted by trained personnel nationally.7 This gap between antenatal care attendance and clinic-based deliveries suggests that there are factors influencing women not to return to the health facilities at the time of delivery, despite information and knowledge about the capacity of skilled personnel to conduct deliveries and identify high-risk
pregnancies and referral cases. A need for qualitative explanatory factors for this led to the study question.
The aim for this explorative study was to identify a variety of individual, community and health facility level factors that may be influencing choice of place of delivery in rural Malawi. It was carried out in the Mangochi district in October to December 2004.
MATERIALS AND METHODS Study area/setting
Malawi is situated in south eastern Africa and has a total population of 11 million people.
This study was conducted in the Mangochi-Lungwena area at the southern shore of Lake Malawi. The Mangochi district has a population of approximately 600,000 and is served by 29 primary health facilities and one district hospital. The district is rural and the people are poor, relying on farming and fishing. Lungwena is one part within the Mangochi district of 26 villages, with approximately 17,000 people. Most of the inhabitants in the southern part of the country are Muslim, and the predominant ethnic group is the Yao tribe.
Mangochi District Hospital is a governmental driven referral hospital, which have a maternity unit with trained midwives and one doctor that supervises the whole hospital. Obstetric emergencies, like caesarean sections, are managed by Medical Officers. In the middle of Lungwena, there is a health centre situated approximately 32 km from the district hospital.8 There are also two private mission hospitals in the Lungwena area. Most of the villages are served by trained traditional birth attendants, and also by traditional healers.
Data collection
Data were collected at Mangochi District Hospital, Lungwena Health Centre and in two villages located eight and one kilometre from the health centre, respectively. Qualitative data collection methods were used. Semi-structured in-depth interviews were conducted with six women that had delivered recently, three of them at home and three at hospital or health centre. They were selected either from the postnatal ward or when attending postnatal check- up on a randomly chosen day during the study period. Two health workers and two traditional birth attendants were included as key informants.
Written signature or oral consent witnessed and signed by a third party was obtained. A semi- structured interview guide was employed. The interview focused on a) the women’s
experiences of and preferences for delivery care and their ideas about barriers to use of professional obstetric care; b) their perceptions and interpretations of danger signs during pregnancy; and c) influencing factors and decision makers in the family and local community.
During the interviews with health workers and traditional birth attendants their ideas of the same issues were probed. An interpreter was used whenever necessary.
In addition to the interviews, non-participating informed observations were conducted, where one of the researchers, with known medical background, observed without taking part in the delivery or antenatal services performed. This took place at the delivery ward, antenatal clinic and postnatal check up at Mangochi District Hospital, and also during antenatal care at Lungwena Health Centre at several randomly chosen points of time. The three authors also participated in focus group discussions on this theme in a few villages. (Conducted in Yao language)
Analysis
All of the interviews were tape recorded and transcribed in full. The transcriptions were analysed in an open way using content analysis, where the narrative qualitative information and the observations were organized and integrated according to emerging themes and concepts.
Ethical approval
The College of Medicine Research and Ethics Committee (COMREC) in Malawi approved the study.
Limitations
The translation may have introduced biases. An additional independent interpreter controlled parts of the recordings/transcripts to prevent this effect. A small number of study participants may not give a representative picture. Still, the information gathered from the different sources showed convincing consistency. The medical background of interviewer and
interpreter may have influenced answers towards a more positive attitude to modern medicine.
The observations were done from an outsider’s point of view, with preferences for modern western medicine.
FINDINGS AND DISCUSSION
The analysis identified that the most important factors influencing women’s choice of place of delivery fall into the following categories: (1) Quality of care, (2) Cultural factors and (3) Availability.
1. Quality of care
Perhaps surprisingly, many of the findings in this study could in some way or another be associated with different aspects of sub optimal care, and therefore to a large extent be attributed to the health care system itself.
a) Communication and attitudes Health education and information given
In this study, the focus was on the women themselves, to reveal their (expressed) reasons for their choices. However, the role of antenatal care (ANC) in educating and advising women is evident. Therefore, some questions about antenatal care were included.
All of the women involved had attended antenatal care during all their pregnancies, like the great majority of pregnant women in Malawi do. The reasons given for attending antenatal care were that they wanted to know the growth and progress of the pregnancy, to obtain drugs and immunisations, to know if everything was normal, and to have any problems diagnosed and treated. (One woman also mentioned that attending ANC would give her a good delivery, and would prevent her from delivering before term.)
Unexpectedly, attending antenatal clinics two or more different places was not unusual, also according to the nurse at the health centre who witnessed this often. Her opinion was that this practise was due to the pregnant women’s fear of refusal if they did not have “the right“
ANC-card/ “ticket”, since different types of care providers (private, NGO, government) issue different cards. These women seemed to believe that they for instance needed one “ticket”
from the health centre, if that was where they intended to deliver, one from the TBA, and one from the referral hospital. One of the TBAs also reported that she usually encouraged the pregnant women to go to ANC at the health centre as well as coming to her for controls.
During ANC observations, it was noted that very little time was spent on each patient. One nurse on duty often performed the full examination on all the pregnant women coming for care. There might be more than 70 patients daily. Hence, in the context of Mangochi, ANC did perform the essential routine procedures, but left little time for individualized health education or counselling. Place of delivery as an individual client plan, was not discussed.
The health workers role in advising the women on where to deliver is central in this context.
Some of the interviewees also stated that nobody had talked to them about plans for delivery.
One of the health workers narrated:
“We encourage them to come to the hospital, but if they want to go to the TBA (traditional birth attendant) it’s up to them, (…) all of them cannot deliver at hospital”
The reason for this being lack of space, and in that case a problem joining the long line of equipment and facility scarcities in this poor country. It is possible to deliver at the health centre in the village as well, but they can only handle normal deliveries, and referral from those facilities may face serious delays.
Important issues raised by other researchers are the function, value and quality of antenatal care. 9 Is the contribution in educating pregnant women worth the costs and most importantly, is it capable of recognising those women at risk and through that strategy reduce maternal mortality and morbidity?
In general, pregnant women in Malawi tend to come too late for the first visit, with only 9%
attending during the first trimester. Furthermore, on average, women only attend antenatal clinics three times7 while the recommended number for uncomplicated pregnancies is a minimum of four visits.
It is important to question how the rumours of the possibility of refusal for not having the right “ticket” have been generated. Is it just gossip, built on misunderstandings, uncertainty and fear of unfamiliar environment, or is there some truth in them, after all? At some level, at least there is a communication problem. Another interesting point here is whether the ANC attendance might be overestimated, if this is a widespread phenomenon. For example, one woman attending ANC at three different places may actually be registered as three women.
This also demonstrates lack of communication and coordination between various types of health care delivery.
Client-provider interaction
Approximately 4000 deliveries are conducted at Mangochi District Hospital every year. There are a total number of 80 beds for maternity admissions. The labour ward consisted of one room with seven beds. Nine nurses/midwives and two Clinical Officers were employed at this ward, with three nurses working the day shift and two during the night.
Privacy was difficult to achieve in the delivery room. Some of the beds did not have curtains, and many people had access to the ward. The women in labour waiting for a bed sat on the floor. Confidentiality, thus, was not taken care of.
Sterile delivery packs were present, but they lacked pairs of scissors, so the women had to bring razorblades from home for cutting the cord. They also had to bring towels and blankets for the newborn and black plastic sheets for themselves. Some infection prevention measures were taken, like the use of gloves, decontamination of beds etc, and plastic bins for different kinds of waste stood by the beds. The hand washing, however, was not always a routine.
Some of the health workers wore casual clothing, which might have caused some confusion, since the ward was so busy.
Generally, the staff at the ward talked very little to the patients. Some women in labour had to wait very long for an examination and were often left more or less to themselves. Some even delivered without any assistance, supervision or attention at all.
Deaths of newborns occurred that probably could have been avoided. For instance, one infant died because the cord was not clamped well, and nobody looked after him for a very long time, until the blood loss was too significant and it was too late. Data, like blood pressure, that had not been measured, were written in the records, without caring for the accuracy of information.
Although some complaints were put forward, not many of the women interviewed complained about the treatment they had received, at least not in direct words. But some of them made expressions like preferring to be assisted by a male nurse,
- “Because he is better and kinder than a female nurse”
indicating that their experiences at least with female nurses were not entirely positive.
There are several reasons for this lack of criticism. Possible explanations for not complaining are fear of reprimands, loyalty towards the authorities, which is common in Malawi, or simply that they have never known any other reality; the existing level of quality is what they expect.
The health workers themselves on the other hand, felt that the patients did not understand the pressure on them, and that they were expecting too much.
Most of the people living in this area are very poor, and the socio-economic gap between the well-educated health workers and the patients was naturally prominent. This complicated the communication because of differing perceptions and interpretations of important matters. This challenge becomes a problem when differences turn into distance and disregard. The poor conditions observed at the labour ward turned out to be among the most important findings in this study, illustrating a main challenge for the health care system in Malawi.
The reasons for the increase in maternal mortality in Malawi over the past decade can be attributed not only to the impact of the AIDS epidemic but also to substandard obstetric care in hospitals, the latter being associated with half of the maternal deaths in the Southern region of the country.10
A woman’s general impression of pregnancy and labour is to a large extent influenced by her experiences in the delivery situation.11 Women’s delay in medical care seeking or non- compliance with referral care instructions is presumably partly a result of previous negative experience with the health care system.12 Along with distance and lack of money, rudeness of health workers was in Malawi Obstetric Quality of Care Assessment 13 found to be one of the major constraints to accessing maternal health services.
Nevertheless, the reality is that the health workers are well educated and have resources that are beyond what the poor pregnant women can ever expect to achieve. However, they may have other aspirations for their life than what is currently available and may use their relative indifference as a coping mechanism for fatigue and helplessness in the real situation in which they work with little supervision, poor feedback, and few prospects for any improvement in the near future. The Malawi Obstetric Quality of Care Assessment found that 60% of the health workers in the country were unsatisfied with their salaries, accommodations or other aspects of their working conditions.
Cooperation within the health care system
The TBAs are conducting a considerable proportion of the deliveries in the district (33%).7 However, the professional health workers did not seem to have a lot of trust in them: “ She (the TBA) is maybe trained, but she does things as if she doesn’t have any knowledge about the deliveries.”
In the villages on the other hand, the TBAs gave the impression of being prominent and trusted: "People know me as a TBA, and they know I can assist them”. They also seemed to
have quite a lot of power, for instance through involving the husbands of their clients in their work.
The TBAs interviewed were trained, and knew that they were supposed to refer clients when facing problems. According to the health workers though, this was not always practiced. The TBAs, in contrast, told stories about not being met at the health centre when seeking advice.
Whoever is right or wrong, the role of the TBAs and the value of training them in relation to good outcomes of deliveries is disputed, especially because there are some things they cannot do, like handling an instrumental delivery. 14
b) Facilities, staff and equipment
Also the communication between district and referral hospital was sometimes substandard, due to unreliable wireless messages, lack of transport and bad road conditions, just to name a few. Seeking care at a facility that could not provide the required care thus delayed the medical intervention rather than being able to assist the women in getting optimal treatment quickly.15
Lack of staff is a significant problem, as illustrated by one woman’s statement: “I went to ...
health centre, and the health worker was not there”. Medical Assistants and Clinical Officers were doing most of the work traditionally done by medical doctors, trying to meet the overwhelming needs. Although there were more nurses than doctors, they were too few and overloaded with work. At health centre level there are no trained doctors, and the few doctors at district level continually work as administrators instead of as clinicians. This is a well-
known problem all over the country, that cause a negative effect on the delivery of obstetric and other services.13
The facilities were not properly equipped. Shortage of drugs was common. This situation, with overwhelming workload, lack of manpower and equipment scarcities, leads to frustration, fatigue, sub optimal quality of care, and unsatisfied patients.
2. Cultural factors
Cultural factors influence on health care-seeking behaviour in all communities. In this study, especially the influence from decision-makers, perception of danger signs and traditional views on pregnancy and motherhood were found to be important for the choice of place of delivery.
a) Influence from decision-makers
The women narrated that if they could choose, they would prefer to deliver in a health institution assisted by a professionally skilled person. Reasons given for choosing to deliver at health facilities were almost the same as for attending antenatal care, namely solving problems, or as a safety measure to ensure a positive outcome.
Nevertheless, this decision was not simply their own and the influence from other members of the household seemed strong. For example, older women in the family, like mothers, grandmothers and mothers-in-law, but also the husband, were mentioned as decision-makers.
When it came to those influenced by family to deliver at home, reasons revealed were that the family needed to observe the labour and to be in charge. If someone from the family did not witness the delivery, they might reject the child: “If a woman delivers at hospital, the
husband says: -that’s not my baby.” A childbirth is not merely a medical event, it is a social process and a ritual.
b) Perceptions of danger signs
The most often mentioned predictors for an adverse pregnancy outcome in this study were bleeding, fever/malaria (the same word in the local language), oedema (“swollen feet and body”) and vomiting.
The women’s perceptions of some medically defined risk factors were also probed for. For instance, older multigravidas were perceived as weak; they might bleed and even die. The problem with many, and not very well cared for, children at home was also brought up. The very young and primigravida were seen as at risk of dying, and so was the baby. Becoming ill in other ways might also be the consequence of such deliveries. The association between fever and mosquito bites was well known, so was the importance of sleeping under mosquito nets for pregnant women. They were also familiar with the connection between anaemia (“lack of blood”) and diet. Many of the other symptoms, such as swollen feet, general pains, dizziness and headache were seen as a consequence of “lack of blood”. However, controversy existed around the preferable treatment. Thus, the women knew about danger signs, but did not necessarily act upon them promptly.
Expert-defined risk factors are in other studies found to have little influence on a woman’s decision to seek hospital care, due to different perceptions and interpretations of danger signs.16 The importance of this for pregnant women’s health care seeking behaviour in Malawi is a vital issue that needs further investigation. Certainly, it is essential to keep in mind in future evaluation and implementation of maternity care.
c) Traditional views on delivery and motherhood
In pregnancy, delivery and motherhood matters it was the older women in the community (e.g., mothers, grandmothers and mothers-in-law) who were trusted and perceived knowledgeable, and their advices were listened to. One of the nurses narrated: “The grandmother, she just tells the granddaughter: - don’t worry, deliver here so that I can take care of you. So the girl listens more to the grandmother than she remembers what was said at the clinic.”
Fear of having an operation was mentioned as reason for not going for a hospital delivery.
Caesarean sections were in general considered as negative because “a woman is born to deliver vaginally.” It was also mentioned that it was dangerous and that the reason for having an operation was being lazy. In addition, “those who have an operation, they get fluids and blood which is a threat to our culture.” The longer hospital stay that also comes with an operative delivery was a problem because of time costs.
The attendance for postnatal check up in Malawi is even lower than the hospital deliveries.
(24%)7 Instead, the women preferred to “take herbs at home,” according to one of the interviewees. If nothing was wrong with neither mother nor child, they did not see the point of coming to the clinic. Pregnancy and motherhood are in many cultures perceived to be a natural phenomenon, not requiring intervention unless something is clearly wrong.17
3. Availability
With half of the people in this area living more than 5 km from the nearest hospital, access to health care is sub optimal. In addition to distance, financial and time costs were reducing the availability of maternity health services.
a) Distance and transport
As expected, distance and transport were considered as major constraints by the interviewees.
One of the health workers explained:
“Most of the women they stay very far from the hospital. Some places there are no labour (i.e.
means of motorized) transport, … they use bicycles, so it’s a problem when the mother starts the labour there … That’s why most of the women just deliver at the TBA’s. They can’t reach the hospital.”
Not only the lack of infrastructure and means of transportation, but also the considerable travelling time was given as reason for not reaching hospital. (“Some have a quick labour, so that they cannot walk to a hospital.”) In the rainy season, the road conditions are even worse, and at any time of year it is difficult and dangerous to travel by night. One can imagine walking or bicycling 5-10 km while in labour, at night. It is simply a task not possible to undertake.
In rural Malawi, assistance at delivery by a trained health worker is found to be less likely as the distance to the health facility increases.18 Other researchers have also found distance to be one of the most important determinants, in the decision of not seeking modern health care even when needed.9
b) Financial costs
Lack of money was given as reason for not using the health facilities in an area also served by a private mission based health organization that charged for services. No money may also be a
problem in reaching free maternity care, for example due to transportation costs. And even the governmental driven hospitals expected the women to bring certain equipment as mentioned earlier (razor blades, plastic sheet, blankets). Furthermore, many of the TBAs charge a little fee too, so although contributing, financial costs are probably seldom the only factor that influence place of delivery.
Kowalewski et al 19 found that even though cash payment for the services may be an acute barrier to using maternity care services, in cases where the family considers hospital treatment necessary, efforts are made to cover the costs, and the majority finds the means to afford the admission.
c) Time costs
Time costs are meant opportunity costs of foregone wages, time spent on travel, waiting and treatment; and time spent by caretaker/ accompanying person.
For some of the women, being away from home for several days was a major concern. Who was going to take care of their duties at home? Who was going to look after their other children? Other researchers have concluded that time costs are constantly higher than financial costs. 19
CONCLUSIONS
As Kowalewski et al2 we conclude that the barriers to reaching professional delivery care go far beyond distance and costs. Moreover, they can at least partly, be attributed to the health care system itself. Furthermore, the sub optimal quality of care observed is not only caused by lack of staff, equipment and drugs. In addition to the long-term needs for general poverty
reduction, infrastructure improvements, more educated personnel and better facilities and equipment, this study shows that other kinds of obstacles are equally important and need to be emphasised: An enhanced awareness of the rural pregnant women’s psychological vulnerability and insecurity is necessary, and better education and training of health workers in interpersonal skills and attitudes is required.
Different perceptions and interpretations of danger signs during pregnancy amongst health personnel and the local community and traditional views on pregnancy and motherhood, are important cultural factors influencing health care-seeking behaviour that must be kept in mind and made a subject of further research. Allowing a family member or a friend of the woman to accompany her during labour might be a possible intervention to overcome the cultural need of the family to observe the delivery, as well as reducing the barrier of unfamiliar environment that may intimidate or cause anxiety.
It would probably be valuable for increasing not only knowledge, but also trust, to spend a few more minutes than is currently done on each patient during ANC. The time was often too scarce for appropriate counselling, health education and discussion of plans for delivery. One or two focused meeting points may be even more efficient than four to five ritualistic ones.
Influence from decision-makers was also found to be an important obstacle for institutional deliveries. Perhaps including the husband or a family member in parts of the antenatal counselling is an intervention worth exploring.
Even though this explorative study has revealed areas that need further investigation, it is still plausible to draw one main conclusion from the findings; In rural Malawi, a more individualized professional maternity care is needed, that takes into account the particular
woman’s needs, preferences and insecurity, and that meets her as the service-minded institution it is meant to be.
ACKNOWLEDGEMENTS
We would like to acknowledge “MEDECOs stiftelse” and “Skipsreder Tom Wilhelmsens stiftelse” for providing financial support. We also wish to acknowledge NUFU and College of Medicine, Agnes Chimbiri, Jan J. Hofman and Sophie Phiri for the assistance they provided, the staff and leadership at MDH and LHC, and the leadership of the villages we visited.
Especially, we want to express our thanks to the women, health workers and TBAs who willingly gave us their time.
REFERENCES
1. AbouZahr C and Wardlaw T. Maternal mortality in 2000: Estimates developed by WHO, UNICEF and UNFPA, unedited version. UNICEF Statistics, 2003
http://www.childinfo.org/maternal_mortality_in_2000.pdf.
2. Kowalewski M, Jahn A and Kimatta S.S. Why do at-risk mothers fail to reach referral level? Barriers beyond distance and cost. African Journal of Reproductive Health 2000; 4(1): 100-9.
3. Millennium Development Goals [http://ddp-ext.worldbank.org/ext/MDG/gdmis.do]
4. Government of Malawi, Malawi Demographic and Health Survey 1992, 2000.
National statistic Office 1994, 2000.
5. Buttiëns H, Marchal B and De Brouwere V. Editorial: Skilled attendance at childbirth:
let us go behind the rhetorics. Tropical Medicine & International Health 2004; 9(6):
653-654.
6. Ronsmans C, Etard J.F, Walraven G, Høj L, Dumont A, de Bernis L and Kodio B.
Maternal mortality and access to obstetric services in West Africa. Tropical Medicine and International Health 2003; 8 (10): 940-948.
7. Health Management Information Bulletin, Malawi, December 2003.
8. Espo M. Infant mortality and its underlying determinants in rural Malawi. PhD thesis, University of Tampere, Tampere 2002, 49-52
9. Kulmala T. Maternal Health and pregnancy outcomes in rural Malawi. PhD thesis, University of Tampere, Tampere 2000.
10. Safe Motherhood Project. The Confidential Enquiries into Institutional maternal deaths in the Southern Region of Malawi, 2001
11. Bergstrøm S. Förlossningsvård i U-land. Reklam & Katalogtryck, Uppsala 1988.
12. Lawson J.B, Harrison K.A and Bergstrøm S. Maternity care in developing countries, RCOG Press, 2001
13. Reproductive Health Unit, Ministry of Health and Population in Malawi, WHO.
Malawi Obstetric Quality of Care Assessment, Draft report, 2003.
14. Jahn A, Carvalho I and Kalinga M.J. Evaluating traditional midwife training programs: lessons learned from Tanzania. International Journal of Gynecology &
Obstetrics 2001, 73: 277-278.
15. Cham M, Sundby J, and Vangen S. Maternal mortality in the rural Gambia, a qualitative study on access to emergency obstetric care. Reproductive Health 2005;
2(3). (Online publication: http://www.reproductive-health-journal.com/content/2/1/3 )
16. Jahn A, Kowalewski M and Kimatta S.S. Obstetric care in southern Tanzania: does it reach those in need? Tropical Medicine and International Health 1998; 3(11): 926- 932.
17. Caldwell B.K. Identification of unmet needs. Why women continue to die from childbirth in Dhaka, Bangladesh. Reproductive Health, Unmet needs and Poverty.
Issues of access and quality of services. Lerner S and Vilquin E (eds.), Paris, Cicred, 2005.
18. Van der Broek N.R, White S.A, Ntonya C, Ngwale M, Cullinan T.R, Molyneux M.E and Neilson J.P. Reproductive health in rural Malawi: a population-based survey.
BJOG: an International Journal of Obstetrics and Gynecology 2003; 110: 902-908.
19. Kowalewski M, Mujinja P and Jahn A. Can mothers afford maternal health costs?
User costs of maternity services in rural Tanzania. Afr J Reprod Health 2002; 6(1): 65- 73
